Release 392 (3 Aug 2022)#
General#
Add support for dynamic filtering when task-based fault-tolerant execution is enabled. (#9935)
Add support for correlated sub-queries in
DELETEqueries. (#9447)Fix potential query failure in certain complex queries with multiple joins and aggregations. (#13315)
JDBC driver#
Add the
assumeLiteralUnderscoreInMetadataCallsForNonConformingClientsconfiguration property as a replacement forassumeLiteralNamesInMetadataCallsForNonConformingClients, which is deprecated and planned to be removed in a future release. (#12761)
ClickHouse connector#
Report the total time spent reading data from the data source. (#13132)
Delta Lake connector#
Add support for using a randomized location when creating a table, so that future table renames or drops do not interfere with new tables created with the same name. This can be disabled by setting the
delta.unique-table-locationconfiguration property to false. (#12980)Add
delta.metadata.live-files.cache-ttlconfiguration property for the caching duration of active data files. (#13316)Retain metadata properties and column metadata after schema changes. (#13368, #13418)
Prevent writing to a table with
NOT NULLor column invariants columns. (#13353)Fix incorrect min and max column statistics when writing
NULLvalues. (#13389)
Druid connector#
Add support for
timestamp(p)predicate pushdown. (#8404)Report the total time spent reading data from the data source. (#13132)
Change mapping for the Druid
floattype to the Trinorealtype instead of thedoubletype. (#13412)
Hive connector#
Add support for short timezone IDs when translating Hive views. For example,
JSTnow works as an alias forAsia/Tokyo. (#13179)Add support for Amazon S3 Select pushdown for JSON files. (#13354)
Iceberg connector#
Add support for hidden
$file_modified_timecolumns. (#13082)Add support for the Avro file format. (#12125)
Add support for filtering splits based on
$pathcolumn predicates. (#12785)Improve query performance for tables with updated or deleted rows. (#13092)
Improve performance of the
expire_snapshotscommand for tables with many snapshots. (#13399)Use unique table locations by default. This can be disabled by setting the
iceberg.unique-table-locationconfiguration property to false. (#12941)Use the correct table schema when reading a past version of a table. (#12786)
Return the
$pathcolumn without encoding when the path contains double slashes on S3. (#13012)Fix failure when inserting into a Parquet table with columns that have quotation marks in their names. (#13074)
MariaDB connector#
Report the total time spent reading data from the data source. (#13132)
MySQL connector#
Report the total time spent reading data from the data source. (#13132)
Change mapping for the MySQL
enumtype to the Trinovarchartype instead of thechartype. (#13303)Fix failure when reading table statistics while the
information_schema.column_statisticstable doesn’t exist. (#13323)
Oracle connector#
Report the total time spent reading data from the data source. (#13132)
Phoenix connector#
Report the total time spent reading data from the data source. (#13132)
Pinot connector#
Redact the values of
pinot.grpc.tls.keystore-passwordandpinot.grpc.tls.truststore-passwordin the server log. (#13422)
PostgreSQL connector#
Report the total time spent reading data from the data source. (#13132)
Improve performance of queries with an
INexpression within a complex expression. (#13136)
Redshift connector#
Report the total time spent reading data from the data source. (#13132)
SingleStore (MemSQL) connector#
Report the total time spent reading data from the data source. (#13132)
SQL Server connector#
Report the total time spent reading data from the data source. (#13132)